Arising from: Marks JB (2008) How do detemir and glargine compare when added to oral agents in insulin-naïve patients with type 2 diabetes mellitus? Nat Clin Pract Endocrinol Metab 4: 490–491 doi:10.1038/ncpendmet0900

As an original author, I write in reply to Dr Marks' assessment1 of our comparative trial of detemir or glargine added to oral antidiabetic therapy in type 2 diabetes.2 I am flattered our study was thought worthy of comment, and agree it is important to understand the relative clinical profiles of these basal insulins.

Dr Marks correctly points out some limitations in our study design. In particular, she draws attention to the unequal dosing algorithms. While patients randomised to glargine were dosed once daily in the evening, patients randomised to detemir could add a second morning dose; these dosing approaches reflecting product licensing at the time.

Since then, we have learnt that the mean time-action profile of detemir is well suited to once-daily dosing3,4 and the product license has been adjusted accordingly. Thus, as Dr Marks points out, the two analogues should ideally be compared using equal dosing algorithms.

Further evidence for this was provided by our trial. The criteria for changing from once-daily to twice-daily detemir were blood glucose values of <7.0mmol/l (pre-breakfast) and >7.0mmol/l (pre-dinner), and/or hypoglycemia limiting once-daily titration. The split doses were then titrated to blood glucose targets of <6.1mmol/l (pre-breakfast and pre-dinner). When the blood glucose diaries of all patients were retrospectively assessed, it was apparent that had these criteria been applied equally and strictly, then 87% of glargine-treated and 81% of detemir-treated patients would have switched to twice-daily dosing. This illustrates the problem of an apparent waning of effect over 24 hours with both insulins5 and raises potentially important issues concerning pre-dinner glucose levels. How important is pre-dinner glycemia? What are appropriate targets? How best can it be controlled?

I do not propose to discuss these issues here; suffice to say that twice-daily dosing of basal insulin may not be the best approach for pre-dinner glucose control, the reason relating to another finding Dr Marks highlights: a relative increase in the basal dose. Dr Marks wrongly states that patients completing on once-daily detemir had a mean dose of 0.78 U/kg; the correct value was 0.52 U/kg. However, she correctly notes how twice-daily dosing markedly increased this to an average overall dose of 0.78 U/kg. That twice-daily administration leads to dose escalation without proportional gains in glycemic control has been documented for both detemir and glargine, and the mechanisms explained by deVries and colleagues.5

In conclusion, I feel the inferences that our study suggested glargine to be more cost effective (due to lower dose requirements) and to require fewer injections are too simplistic. Although not reported, our study suggested an equal proportion of patients would have required twice-daily glargine given equal algorithms, while the dose discrepancy that emerged is likely to have been, to at least some extent, an artifact of the different dosing algorithms. I do agree, however, that an equal comparison of detemir and glargine is needed, and am pleased to report that Novo Nordisk has recently initiated such a study.